“…Similarly, the potential for red muscle endothermy in L. imperialis was explored by quantifying the longitudinal and transverse distributions of the red muscle and comparing them to those of species known to be ectothermic or regionally endothermic with respect to their red muscle. Literature values were used in quantitative comparison to L. imperialis with respect to the (1) longitudinal distribution of the red muscle as a proportion of total body cross‐sectional area (Dickson et al ., 2000; Graham et al ., 1983; He, 1986; Johnston et al ., 1988; Johnston & Camm, 1987; Malik et al ., 2020; Zhang et al ., 1996), (2) relative red muscle area compared to that at 50% fork length (Bernal et al ., 2003, 2010; Bernvi, 2016; Ellerby et al ., 2000; He, 1986; Sepulveda et al ., 2005), (3) relative distance between the vertebral column and red muscle compared to that at 50% fork length (Perry et al ., 2007), (4) relative distance between the red muscle and body perimeter compared to total body width at 45%–50% fork length (Dickson et al ., 2000; Perry et al ., 2007; Stoehr et al ., 2020) and (5) normalized perimeter index (perimeter of a circle divided by that of the focal shape of equal area, 0 < x ≤1) of the red muscle cross‐sectional shape at 45%–50% fork length (Bernvi, 2016; Dickson et al ., 2000; Malik et al ., 2020; Perry et al ., 2007; Sepulveda et al ., 2005; Stoehr et al ., 2020). Furthermore, generalized additive mixed models (GAMMs) were used to predict across‐endotherm and across‐ectotherm estimates of the longitudinal distribution of red muscle as a proportion of total body cross‐sectional area and relative red muscle area compared to that at 50% fork length (accounting for shape constraint in the latter metric); these GAMMs predicted red muscle metrics as a function of fork length with a global, across‐species smoother and a random effect for individual species ( sensu Pedersen et al ., 2019).…”